Structure of the RGS-like domain from PDZ-RhoGEF: linking heterotrimeric g protein-coupled signaling to Rho GTPases

BACKGROUND

The multidomain PDZ-RhoGEF is one of many known guanine nucleotide exchange factors that upregulate Rho GTPases. PDZ-RhoGEF and related family members play a critical role in a molecular signaling pathway from heterotrimeric G protein-coupled receptors to Rho proteins. A approximately 200 residue RGS-like (RGSL) domain in PDZ-RhoGEF and its homologs is responsible for the direct association with Galpha12/13 proteins. To better understand structure-function relationships, we initiated crystallographic studies of the RGSL domain from human PDZ-RhoGEF.

RESULTS

A recombinant construct of the RGSL domain was expressed in Escherichia coli and purified, but it did not crystallize. Alternative constructs were designed based on a novel strategy of targeting lysine and glutamic acid residues for mutagenesis to alanine. A triple-point mutant functionally identical to the wild-type protein was crystallized, and its structure was determined by the MAD method using Se-methionine (Se-Met) incorporation. A molecular model of the RGSL domain was refined at 2.2 A resolution, revealing an all-helical tertiary fold with the mutations located at intermolecular lattice contacts.

CONCLUSIONS

The first nine helices adopt a fold similar to that observed for RGS proteins, although the sequence identity with other such known structures is below 20%. The last three helices are an integral extension of the RGS fold, packing tightly against helices 3 and 4 with multiple hydrophobic interactions. Comparison with RGS proteins suggests features that are likely relevant for interaction with G proteins. Finally, we conclude that the strategy used to produce crystals was beneficial and might be applicable to other proteins resistant to crystallization.

A novel and simple technique to allow detection of the position of the R-waves from intraventricular pressure waveforms: application to the conductance catheter method

A simple and novel technique that utilizes the zero-crossing points of the first time derivative of intra-ventricular pressure (dP/dt) to mark systole, is proposed. Discrete differentiation of the sampled pressure waveform is calculated using a difference equation. Filtration of high-frequency noise in dP/dt is achieved using a low-pass Butterworth filter of order 4 and a cutoff frequency of 10 Hz. The filter is realized digitally using infinite impulse response filter stages. Double filtering of discrete dP/dt is used to eliminate time shifts. The methods are evaluated on data obtained from six large, white, anaesthetised and open chest pigs, instrumented with a conductance catheter. The zero-crossing points of the filtered dP/dt compare very well with the R-waves of the electrocardiogram (ECG) as markers of systole. The mean error is 1.3% of the duration of the heartcycle. Significantly, our results provide a solution to a problem often encountered with multiuse pressure-volume catheters when an ECG signal cannot be obtained. In this situation, the zero-crossing points of dP/dt, rather than the R-waves of the ECG, can be used as a marker of systole, thus enabling the construction of end-systolic pressure-volume relations to assess cardiac contractility.

Expression, purification, and crystallization of the RGS-like domain from the Rho nucleotide exchange factor, PDZ-RhoGEF, using the surface entropy reduction approach

Lsc-homology domains are found in several eukaryotic nucleotide exchange factors which act on Rho-family GTPases. They show limited amino acid sequence similarity to RGS proteins, which down-regulate the cellular signaling by the alpha-subunits of trimeric G-proteins and have been shown to interact with Galpha12 and Galpha13. It is believed that the RGS-like (RGSL) domain constitutes the functional link between G-protein-coupled receptors and cytosolic Rho-GTPases. We report here the expression, purification, and crystallization of the RGSL domain from the PDZ-RhoGEF. To obtain X-ray-grade crystals we have used the recently proposed approach of crystallization by mutational surface entropy reduction, in which selected Lys --> Ala, Glu --> Ala, and/or combined point mutations are introduced into the target protein to reduce the cumulative conformational entropy of surface residues. Of the five mutants that were designed and prepared, the second one tried (K463A, E465A, E466A) yielded crystals suitable for further analysis and diffracted X-rays to 2.8 A resolution on a home source. The crystals exhibit hexagonal symmetry, space group P6(1) 22 or P6(5) 22, with unit cell parameters a = b = 63.1 A, c = 202.1 A, and contain one molecule in the asymmetric unit.

Coronary artery surgery for ischemic heart failure: risks, benefits, and the importance of assessment of myocardial viability

Heart failure and left ventricular dysfunction are common and are most often caused by myocardial ischemia/infarction secondary to occlusive coronary artery disease. Although recent refinements in medical therapy have resulted in improved survival, morbidity and mortality remain high in patients with advanced heart failure. Heart transplantation remains an option for selected patients, and implantable left ventricular assist devices may soon provide another treatment strategy for such patients. However, patients with established postischemic heart failure, significant myocardial viability, and coronary artery anatomy amenable to surgical revascularization can derive significant functional and survival benefit after coronary artery surgery, albeit with an increased perioperative risk. We discuss the role of coronary artery surgery in ischemic heart failure and review the evidence for such an approach.

Conservation of small-airway function by tacrolimus/cyclosporine conversion in the management of bronchiolitis obliterans following lung transplantation

We studied serial lung function in 11 patients with bronchiolitis obliterans syndrome who were treated with tacrolimus conversion following lung or heart-lung transplantation. Our results show that tacrolimus conversion slows the decline of lung function in bronchiolitis obliterans syndrome. The attenuation continues for at least 1 year following conversion.

Clinical and patho-anatomical factors affecting expansion of thoracic aortic aneurysms

OBJECTIVE

To examine the expansion of aneurysmal aortic segments (> or = 35 mm) and to assess the impact of clinical and patho-anatomical factors on aneurysm expansion.

DESIGN

87 consecutive patients (mean age 63.6 years, range 22-84 years) were studied using serial (six month intervals) computed tomographic or magnetic resonance imaging to monitor progression of thoracic aortic aneurysms. Aortic diameter was measured at seven predetermined segments and at the site of maximum aortic dilatation (MAX).

RESULTS

780 segment intervals were identified. The median overall aneurysm expansion rate was 1.43 mm/year. This increased exponentially with incremental aortic diameter (p < 0.01) and varied by anatomical segment (p < 0.05). The presence of intraluminal thrombus (p < 0.01) but not dissection or calcification was associated with accelerated growth. Univariate analysis identified thrombus (p < 0.001), previous stroke (p < 0.002), smoking (p < 0. 01), and peripheral vascular disease (p < 0.05) as factors associated with accelerated growth in MAX. Dissection, wall calcification, and history of hypertension did not affect expansion. beta Blocker treatment was not associated with protection. Multivariate analysis confirmed the positive effect of intraluminal thrombus and previous cerebral ischaemia, and the negative effect of previous aortic surgery on aneurysm growth. These findings translated into a mathematical equation describing exponential aneurysm expansion.

CONCLUSIONS

Aneurysmal thoracic aortic segments expand exponentially according to their initial size and their anatomical position within the aorta. The presence of intraluminal thrombus, atherosclerosis, and smoking history is associated with accelerated growth and may identify a high risk patient group for close surveillance.

Phenotypic correction of feline lipoprotein lipase deficiency by adenoviral gene transfer

Previous studies have revealed that adenovirus-mediated ectopic liver expression of human LPL (huLPL) can efficiently mediate plasma triacylglycerol (TG) catabolism in mice despite its native expression in adipose and muscle tissue. We aimed to explore the feasibility of liver-directed gene transfer and enzyme replacement for human LPL deficiency in a larger, naturally occurring feline animal model of complete LPL deficiency that is remarkably similar in phenotype to the human disorder. A cohort of LPL-deficient (LPL -/-) cats was given an intravenous injection of 8 x 10(9) PFU/kg of a CMV promoter/enhancer-driven, E1/E3-deleted adenoviral (Ad) vector containing a 1.36-kb huLPL cDNA (Ad-LPL) or reporter alkaline phosphatase gene (Ad-AP). After Ad-LPL administration, active, heparin-releasable huLPL was readily detected along with a 10-fold reduction in plasma TGs, disappearance of plasma TG-rich lipoproteins up to day 14, and enhanced clearance of an excess intravenous fat load on day 9. However, antibody against the huLPL protein was detected on day 14 in cats receiving Ad-LPL and adenovirus-specific neutralizing antibody was present 7 days after gene transfer in both cat cohorts. Tissue-specific expression of the huLPL transgene relative to controls was confirmed by RT-PCR. While huLPL expression was evident in the liver, other tissues including spleen and lung expressed huLPL message, in direct correlation with histological evidence of increased Oil red O (ORO)-positive neutral lipid influx. In contrast, intravenous LPL enzyme replacement therapy (ERT) led to rapid disappearance of 9000 mU/kg of active bovine LPL enzyme from the circulation, with t1/2 occurring at <10 min in two LPL-/- cats. Heparin injection 1 hr later released <10% of the original bovine LPL, further indicating its rapid systemic clearance, inactivation, or degradation as well as its ineffectiveness as a viable therapeutic alternative for complete LPL deficiency. Although LPL gene transfer and expression via this first-generation Ad vector was limited by the immune response against both the human LPL protein and adenovirus our results clearly provide a key advance supporting further development of LPL gene therapy as a viable therapeutic option for clinical LPL deficiency.

Replacement of the proximal aorta and aortic valve using a composite bileaflet prosthesis and gelatin-impregnated polyester graft (Carbo-Seal): early results in 143 patients

OBJECTIVE

We report the combined early results from two centers in the United Kingdom using a composite conduit consisting of a bileaflet mechanical valve incorporated into a gelatin-impregnated, ultra-low porosity, woven polyester graft (Carbo-Seal; Sulzer Carbomedics, Inc, Austin, Tex).

METHODS

Between August 1992 and March 1997, 143 patients underwent aortic root replacement with the Carbo-Seal composite prosthesis. The indication for surgery was acute type A dissection in 31 (22%), chronic type A dissection in 9 (6%), ascending aortic aneurysm without dissection in 100 (70%), and false aneurysm of the ascending aorta in 3 (2%). Twenty-seven patients (19%) had undergone previous sternotomy, and 40 (28%) were seen as emergencies. Concomitant procedures were performed in 38 (27%), including 18 aortic arch or hemiarch replacements. Total follow-up is 270 patient-years. Follow-up is 100% complete.

RESULTS

The early (30-day) mortality was 7% (10 patients). Permanent neurologic events occurred in 2%. At a mean follow-up of 23 months, 94% of survivors were in New York Heart Association functional class I. Freedom from reoperation was 97.2% +/- 1.6% (1 standard error [1 SE]) at 12 months and 95.7% +/- 2.2% at 48 months. Including early mortality, survival was 90.1% +/- 2.6% at 12 months and 83.1% +/- 3. 5% at 48 months.

CONCLUSIONS

Aortic root replacement with use of the Carbo-Seal prosthesis can be undertaken with a relatively low early mortality and morbidity. A low reoperation rate and high intermediate-term survival can be expected, but continued follow-up is needed to determine the long-term efficacy of this prosthesis.

Coronary revascularisation for postischaemic heart failure: how myocardial viability affects survival

OBJECTIVE

To assess the impact of revascularisation of viable myocardium on survival in patients with postischaemic heart failure.

METHODS

35 patients (mean (SD) age 58 (7) years) with severe heart failure (New York Heart Association (NYHA) functional class > or = III), mean left ventricular ejection fraction (LVEF) 24 (7)% (range 10-35%), and limited exercise capacity (peak oxygen consumption (VO(2)) 15 (4) ml/kg/min) were studied. 21/35 patients had no angina. Myocardial viability was assessed with quantitative positron emission tomography and the glucose analogue (18)F-fluorodeoxyglucose (FDG) (viable segment = FDG uptake > or = 0.25 micromol/min/g) in all patients before coronary artery bypass grafting. Patients were divided into two groups: group 1, > or = 8 viable dysfunctional segments (mean 12 (2), range 8-15); and group 2, < 8 viable dysfunctional segments (mean 3.5 (3), range 0-7). The two groups were comparable for age, sex, NYHA class, LVEF, and peak VO(2).

RESULTS

Two patients died perioperatively and seven patients died during follow up (mean 33 (14) months). All deaths were from cardiac causes. Kaplan-Meyer survival analysis showed 86% survival for group 1 patients versus 57% for group 2 (p = 0.03). Analysis by Cox proportional hazard model revealed three independent factors for cardiac event free survival: presence of > or = 8 viable segments (p = 0.006); preoperative LVEF (p = 0.002); and patient age (p = 0.01).

CONCLUSION

Revascularisation for postischaemic heart failure can be associated with good survival, which is critically dependent upon the amount of viable myocardium.

Infections, vaccinations, and the risk of childhood leukaemia

A nationwide case-control study was conducted in New Zealand, to test hypotheses about the role of infections in the aetiology of childhood leukaemia. Children aged 0-14 years with leukaemia were matched on age and sex to controls selected from birth records. Case ascertainment was virtually complete and 121 (92%) of 131 eligible case families took part. The participation rate among the 303 first-choice eligible controls was 69%. Home interviews and serological tests were conducted. Adjusted relative risks were estimated by logistic regression. There was an increased risk of leukaemia in relation to reported influenza infection of the child during the first year of life (adjusted odds ratio 6.8, 95% confidence interval 1.8-25.7). This could be a chance finding due to multiple comparisons, and it should be tested elsewhere. Some key variables relevant to Greaves' hypothesis were not associated with B-cell precursor acute lymphoblastic leukaemia (numbers of infections and vaccinations, firstborn status, attendance at preschool groups), although a small effect could not be ruled out with a study of this size. Leukaemia risk was higher among children in poorer social circumstances, and this was true for all eligible children as well as for the participants.

Childhood cancer in New Zealand 1990 to 1993

An epidemiological study of childhood cancer in New Zealand identified 409 children aged 0 to 14 years with malignant neoplasms newly diagnosed between 1990 and 1993 inclusive. The original microscopic material on which the diagnoses were based was reviewed in 398 cases and the neoplasms were allocated into the 12 major groupings and 48 further subcategories of the International Classification of Childhood Cancer (ICCC). The pathology reviewers agreed with group and subcategory classification of the confirmed cancers in all but one case of acute leukemia and three cancers of the central nervous system. Changes were also made in the FAB classification of three cases of acute non-lymphocytic leukemia and in the further subcategorisation of three Hodgkin's lymphomas and ten astrocytomas. The results show a high level of diagnostic accuracy for confirmed childhood neoplasms in that time period. Nine of 15 cases of malignant melanoma notified to the study were not confirmed for various reasons, which included a change in the pathological diagnosis in four cases. Compared with Victoria (Australia), New Zealand has a high incidence rate of lymphomas in boys and an unusual female preponderance of Wilms' tumor cases.

Lipid and lipoprotein analysis of cats with lipoprotein lipase deficiency

BACKGROUND

We have previously described a colony of domestic cats with a naturally occurring mutation in the lipoprotein lipase (LPL) gene. We have now further characterized cats homozygous for LPL deficiency (LPL -/-, homozygotes), and have contrasted these with heterozygotes (LPL +/-) and normal cats (LPL +/+).

MATERIALS AND METHODS

Density gradient ultracentrifugation with subsequent lipid analysis, agarose and polyacrylamide gel electrophoresis was used to examine detailed liproprotein differences between the genotypes. Oral fat loading studies and breast milk fatty acid analysis were also performed to further characterize the phenotypic expression of LPL deficiency in this model system.

RESULTS

Several lipid abnormalities associated with homozygosity for LPL deficiency were evident. Triglyceride-rich lipoprotein-triglycerides (TRL-TG) and cholesterol (TRL-C) were higher (TRL-TG 2.09 +/- 1.14 vs. 0.15 +/- 0.04 mmol L-1, P < 0.001; TRL-C 0.42 +/- 0.30 vs. 0.11 +/- 0.16 mmol L-1, P < 0.05) in male -/- than in male +/+ cats, as was HDL-cholesterol (HDL-C, 1.75 +/- 0.24 vs. 1.41 +/- 0.14 mmol L-1, P < 0.05). LDL-C levels were lower in homozygous cats than in control cats, similar to what is seen in human LPL deficiency. Oral fat loading studies revealed that homozygous cats have a marked reduced ability to clear plasma TGs in terms of peak time (7 h vs. 3 h), peak height (9.36 vs. 1.1 mmol L-1), area under the TG clearance curve (AUC, 280.3 vs. 2.2 h mmol L-1) and time to return to baseline. Fasting lipid and lipoprotein levels were not significantly different between heterozygous and normal cats. However, oral fat loading in heterozygotes revealed an intermediate phenotype (peak of 2.35 mmol L-1 at 5 h, AUC 13.1 h mmol L-1), highlighting the impaired TG clearance in these animals.

CONCLUSION

Thus, LPL deficiency in the cat results in a lipid and lipoprotein phenotype that predominantly parallels human LPL deficiency, further validating the use of these animals in studies on the pathobiology of LPL.

Enhanced lipolysis in normal mice expressing liver-derived human lipoprotein lipase after adenoviral gene transfer

The authors previously demonstrated that the gene for human lipoprotein lipase (hLPL), an enzyme crucial to the breakdown of triglyceride (TG)-rich dietary fats, corrects the hypertriglyceridemia in lipoprotein lipase (LPL)-deficient knockout mice after adenoviral (Ad)-mediated LPL gene transfer. They have now extended their observations to primary cultured mouse hepatocytes and intact animals of normal LPL genotype, and confirm effective overexpression of hLPL from the liver and a sustained TG-lowering effect in plasma over 60 days. A typical first-generation Ad-vector containing the hLPL cDNA (Ad-LPL) resulted in efficient gene transfer into isolated mouse hepatocytes and significant de novo synthesis of active hLPL protein. In this experiment, 5 x 10(9) viral particles (5 x 10(7) pfu) of either Ad-LPL or an Ad-LacZ control vector were injected into CD1 mice of normal LPL genotype. Hepatic expression of hLPL was confirmed at Day 7 postinjection by in situ hybridization and direct measurement of LPL in the liver. This correlated with a total LPL activity (human + mouse) in postheparin plasma (PHP) of 1020.5 standard deviation [SD] 93.6 mU/mL, versus 479.5 SD 129.7 mU/mL (p < 0.001) in Ad-LacZ controls at Day 7. Respective hLPL activity comprised 49% of the total. Significantly raised levels of hLPL protein mass persisted until Day 60. Corresponding plasma TGs decreased to 39% of Ad-LacZ controls at Day 7, and, despite absent hLPL activity from Day 28 on, serum TGs remained significantly lower in Ad-LPL mice up to Day 42. Fast phase liquid chromatography analysis showed a dramatic depletion in TG-rich lipoproteins, mainly very low density lipoproteins (VLDL) and chylomicron fractions. Therefore, Ad-mediated overexpression of hepatic LPL was found to significantly decrease plasma TG levels unrelated to primary LPL deficiency.

Prototypes of intrafamily homicide and serious assault among insanity acquittees

Public concern with societal violence is intensified when persons who have been found not guilty by reason of insanity (NGRI) of having committed a homicide or serious assault are returned to the community. Successful management of such acquittees in the community requires a sophisticated understanding of the person and the illness within the larger context of the violent incident, the family, the community, and the culture. In this article, we present an analysis of psychotic violence within a family context. A qualitative study of 64 subjects who were found NGRI of killing or seriously injuring a family member resulted in four prototypes of intrafamilial homicide/assault: Till Death Us Do Part; Overwhelming Burden, Elimination of the Limit Setter; and Family-Focused Delusional Killing. The prototypes are presented as a model for developing management strategies both for future risk assessment and for successful transition of the insanity acquittee into the community.

Insulin-like growth factor-I prevents development of a vincristine neuropathy in mice

Vincristine is a commonly used antitumor agent whose major dose-limiting side-effect is a mixed sensorimotor neuropathy. To assess whether insulin-like growth factor-I (IGF-I), a neurotrophic agent that supports the survival of motoneurons and enhances regeneration of motor and sensory neurons, could prevent the peripheral neuropathy produced by vincristine, mice were treated with both vincristine (1.7 mg/kg, i.p., 2 x /week) and/or IGF-I (0.3 or 1 mg/kg, s.c. daily) for 10 weeks. In mice treated with vincristine alone, there was evidence of a mixed sensorimotor neuropathy as indicated by changes in behavior, nerve conduction and histology. Caudal nerve conduction velocity was significantly slower in mice treated with vincristine alone as compared with vehicle-treated mice. Vincristine treatment alone also significantly increased hot-plate latencies and reduced gait support and stride length, but not toe spread distances. The effects of vincristine were accompanied by degeneration of sciatic nerve fibers and demyelination, indicating a peripheral neuropathy. IGF-I (1 mg/kg, s.c.) administered to vincristine-treated mice prevented the neurotoxic effects of vincristine as measured by nerve conduction, gait, response to noxious stimuli and nerve histology. At a lower dose of 0.3 mg/kg administered s.c., IGF-I partially ameliorated the neuropathy induced by vincristine as this dose only prevented the change in nerve conduction and hot-plate latencies. IGF-I administered alone had no effect on any of these parameters. These results suggest that IGF-I prevents both motor and sensory components of vincristine neuropathy and may be useful clinically in preventing the neuropathy induced by vincristine treatment.

Correction of hypertriglyceridemia and impaired fat tolerance in lipoprotein lipase-deficient mice by adenovirus-mediated expression of human lipoprotein lipase

Humans homozygous or heterozygous for mutations in the lipoprotein lipase (LPL) gene demonstrate significant disturbances in plasma lipoproteins, including raised triglyceride (TG) and reduced HDL cholesterol levels. In this study we explored the feasibility of adenovirus-mediated gene replacement therapy for LPL deficiency. A total of 5 x 10(9) plaque-forming units (pfu) of an E1/E3-deleted adenovirus expressing either human LPL (Ad-LPL) or the bacterial beta-galactosidase gene (Ad-LacZ) as a control were administered to mice heterozygous for targeted disruption in the LPL gene (n = 57). Peak expression of total postheparin plasma LPL activity was observed at day 7 in Ad-LPL mice versus Ad-LacZ controls (834 +/- 133 vs 313 +/- 89 mU/mL, P < .01), and correlated with human-specific LPL activity (522 +/- 219 mU/mL) and mass (9214 +/- 782 ng/mL), a change that was significant to 14 and 42 days, respectively. At day 7, plasma TGs were significantly reduced relative to Ad-LacZ mice (0.17 +/- 0.07 vs 1.90 +/- 0.89 mmol/L, P < .01) but returned to endogenous levels by day 42. Ectopic liver expression of human LPL was confirmed by in situ hybridization analysis and from raised LPL activity and mass in liver homogenates. Analysis of plasma lipoprotein composition revealed a marked decrease in VLDL-derived TGs. Severely impaired oral and intravenous fat-load tolerance in LPL-deficient mice was subsequently corrected after Ad-LPL administration and closely paralleled that observed in wild-type mice. These findings suggest that liver-targeted adenovirus-mediated LPL gene transfer offers an effective means for transient correction of altered lipoprotein metabolism and impaired fat tolerance due to LPL deficiency.

Diet-induced atherosclerosis in the domestic cat

The domestic cat has not been used in studies of atherosclerosis, with the exception of a single study published in 1970. We have further evaluated the susceptibility of the domestic cat to diet-induced atherosclerosis, the ultimate intent being to discern the atherogenic risk due to lipoprotein lipase deficiency in an affected feline kindred with a phenotype very similar to that of the human form of this condition. We subjected a group of normal domestic cats to a moderately high-fat, cholesterol-enriched diet (30% fat and 3% cholesterol) for a period of 2 to 8 months. Plasma lipid levels were monitored monthly. At the time of killing, all organs and the entire vascular tree were removed, sectioned, processed, and stained with hematoxylin and eosin. The entire vascular tree was also stained with Movat's pentachrome and oil red O (ORO) and assessed semiquantitatively (0 to 5+/5+) and quantitatively (mean intimal area and ORO positivity, mm2). Both blood lipid measurements (total cholesterol, high-density lipoprotein-cholesterol, triglycerides, and low-density lipoprotein-cholesterol) and vessel wall lesion assessment (intimal area, mm2) were statistically elevated (p < 0.05) in the cholesterol-fed cats as compared to those on a normal diet. The highest correlations obtained between blood lipid components and vessel wall measures were the percent increase in triglyceride from base line versus the ORO measurement or foam cell grade (r = 0.86), and percent increase in triglycerides versus the intimal area in the lower abdominal aorta (r = 0.91). Similar relationships were found when the intimal area in the brachiocephalic/subclavian vessels was correlated with the absolute triglyceride values (r = 0.85) or with the percent increase in triglycerides (r = 0.83). Thus, we produced atherosclerotic lesions in the cat within 2 to 4 months on a cholesterol-enriched diet; blood lipid levels were highly correlated with lesional measurements in the vessel wall. This study will provide the basis for evaluation of the susceptibility of New Zealand lipoprotein lipase-deficient cats to diet-induced atherosclerosis.

The accuracy and completeness of childhood cancer registration in New Zealand

The New Zealand Cancer Registry (NZCR) is the main source of data on cancer incidence in New Zealand. The accuracy and completeness of registration of childhood cancers (ages zero to 14 years) were assessed during the conduct of a case-control study. Newly diagnosed children (1990-93) were ascertained from three sources: the NZCR; the Patient Management System (hospital admissions and discharges); and the separate Children's Cancer Registry. Pathology reviews were conducted to verify the diagnoses. Capture-recapture methods were used to assess the completeness of ascertainment. During the four-year period, 409 incident cases of childhood cancer were confirmed, giving an age-standardized incidence rate of 131 per million person-years (95 percent confidence interval = 119-144). The NZCR ascertained 395 (97 percent) of these children. In addition, the NZCR notified us of 43 other 'childhood cancer' registrations which were not confirmed as new cases of childhood cancer according to our eligibility criteria. The main reasons for these were coding errors (20 registrations), duplicates (seven), and a change in the pathological diagnosis as a result of the pathology review (seven). The capture-recapture estimate of the total number of incident cases was 410. Overall, the NZCR had good completeness for childhood cancers, but the number of unconfirmed registrations was larger than expected.